Scoppola, B., D. Boccaletti, M. Bevis, E. Carminati and C. Doglioni

  • View
    212

  • Download
    0

Embed Size (px)

Text of Scoppola, B., D. Boccaletti, M. Bevis, E. Carminati and C. Doglioni

  • The Westward Drift of the Lithosphere: A rotational drag?

    B. ScoppolaDipartimento di Matematica, Universita Tor Vergata Roma, Italy

    D. BoccalettiDipartimento di Matematica, Universita di Roma La Sapienza, Italy

    M. BevisDepartment Civil Environmental Engineering - Geodetic Science, Ohio State University, USA

    E. CarminatiC. Doglioni

    Dipartimento di Scienze della Terra, Universita di Roma La Sapienza, Italy

    June 20, 2005

    Abstract

    Net westward rotation of the lithosphere relative to the underlying mantle is a contro-versial phenomenon first attributed to tidal effects, and later to the dynamics of mantleconvection. In spite of a number of independent geological and geophysical arguments forwestward tectonic drift, this phenomenon has received little recent attention. We suggestthat this differential rotation is a combined effect of three processes: 1) Tidal torques act onthe lithosphere generating a westerly directed torque decelerating the Earths spin; 2) Thedownwelling of the denser material toward the bottom of the mantle and in the core slightlydecreases the moment of inertia and speeds up the Earths rotation, only partly counter-balancing the tidal drag; 3) Thin (3 30 km) layers of very low viscosity hydrate channelsoccur in the asthenosphere. It is suggested that shear heating and the mechanical fatigueself-perpetuate one or more channels of this kind which provide the necessary decouplingzone of the lithosphere.

    Keywords: Earths rotation, westward drift, lithosphere, asthenosphere viscosity, decou-pling

    1 INTRODUCTION

    Since the westward drift of the American blocks described by Wegener (1915), there have beena number of papers proposing a global or net westward drift of the lithosphere relative to themantle (Rittmann, 1942; Le Pichon, 1968; Bostrom, 1971). This net rotation is indicated byindependent kinematic observations such as plate motion within the hotspot reference frame (Ri-card et al., 1991; OConnell et al., 1991; Gordon, 1995; Gripp and Gordon, 2002), plate motion

    1

  • relative to Antarctica (Le Pichon, 1968; Knopoff and Leeds, 1972) and geological asymmetries(Doglioni, 1993).

    Tidal or Earth rotation effects were invoked to explain this westward drift (Bostrom, 1971;Knopoff and Leeds, 1972; Moore, 1973). Jordan (1974) and Jeffreys (1975) attacked the theoreti-cal basis of these tidal drag mechanisms, and the model was abandoned. The notion that Earthsrotation influences plate tectonics has been generally discounted due to the requirement of theconservation of the angular momentum of the Earth-Moon system considered as an isolatedsystem. However there is a body of evidence suggesting an astronomical tuning of plate tecton-ics, such as the distribution of plate velocity and seismicity, which tend to decrease toward theEarths poles (DeMets et al., 1990; Heflin et al., 2004). Transform faults are longer in the equa-torial zones, and mantle thermal minima are also concentrated around the equator suggesting amigration of cooler and heavier material at low latitudes due to centrifugal mass redistribution(Bonatti, 1996). The pole-fleeing force is an example of the rotational component acting onplates, particularly on an oblate planet (Eotvos, 1913; Caputo, 1986a; Gasperini, 1993). Polarwander is also influenced by initiation of subduction zones or internal mass redistributions inthe mantle (e.g., Spada et al., 1992).

    As Jordan (1974) noted, the idea of tidal drag as the driving mechanism for plate tectonics isparticularly intriguing (e.g., Bostrom, 1971; Moore, 1973) because it is energetically feasible. Infact the dissipation of energy by tidal friction is slightly larger (1.6 1019 J/yr) than the energyreleased by tectonic activity (1.3 1019 J/yr), e.g., Denis et al., 2002).

    However, Jordan (1974), and later Ranalli (2000) discarded the Earths rotation as thecause of the westward drift, claiming that the viscosity necessary to allow decoupling betweenlithosphere and mantle should be about 1011 Pa s in the intervening asthenosphere. This valueis too low when compared with the present day estimates of the asthenosphere viscosity, rangingbetween 1017 1020 Pa s (Anderson, 1989; Pollitz et al., 1998; Fjeldskaar, 1994; Giunchi et al.,1997; Piersanti, 1999). Therefore the viscosity of the asthenosphere is crucial for understandingthe nature of the westward drift of the lithosphere.

    In this paper we attempt to revitalize the idea of an astronomical origin of the westwarddrift, including new ingredients to the Jordan (1974) model, such as the non-linear rheologyof the mantle, the mechanical fatigue and the irreversible downwelling of the heavier rocks inthe mantle. Moreover we hypothesize the presence of a ultra low-viscosity channel within theasthenosphere (Fig. 1).

    2 WESTWARD DRIFT

    There are a number of open basic questions regarding the westward drift, including: What isits real speed? What is generating it? Does it affect the entire lithosphere or it is rather onlya mean value, with most of the lithosphere moving west, but part of it still moving east inthe opposite direction relative to the mantle? Ricard et al. (1991) proposed that the westwarddrift is only a mean value due to the lower asthenospheric viscosity at the base of the Pacificplate, but geological and geophysical signatures of subduction and rift zones rather show aglobal signature, supporting a global relative eastward motion of the mantle relative to thelithosphere (Fig. 1). Gripp and Gordon (2002) computed an average westward speed of thelithosphere relative to the asthenosphere of up to about 49 mm/yr, using the hotspot referenceframe.

    2

  • It is crucial to detect whether hotspots are fixed relative to each other (Molnar and Stock,1987; Steinberger, 2002) in order to have a reliable hotspot reference frame and to compute thewestward drift of the lithosphere. Norton (2000) grouped hotspots into three main families thathave very little internal relative motion (Pacific, Indo-Atlantic and Iceland). In his analysis,Pacific hotspots are nearly fixed relative to each other during the last 80 Ma.

    We do not have yet a reliable constraint on the source depth of the hot spots (deep mantle orasthenosphere) but, whatever this depth, hotspots indicate relative motion between lithosphereand the underlying mantle.

    If hotspots have their source in the asthenosphere (Bonatti, 1990; Doglioni et al., 2005), andwe disregard those hotspots located along plate margins do not constitute a reliable referenceframe because they are moving relative to each other, then the net rotation of the lithosphererises to 90 mm/yr. The only assumption made is that the Pacific hotspot tracks are fixed relativeto each other (Norton, 2000; Gripp and Gordon 2002) and that they parallel the motion of theunderlying mantle relative to the lithosphere.

    The most obvious place for decoupling to occur is the asthenosphere where the lowest mantleviscosity values are believed to occur. Kennedy et al. (2002) have shown how mantle xeno-liths record a shear possibly achieved at the lithosphere-asthenosphere interface. This supportsthe notion of a flow in the upper mantle and some decoupling at the base of the lithosphere(Russo and Silver, 1996; Doglioni et al., 1999; Bokelmann and Silver, 2000). A significant radialanisotropy, with horizontally polarized shear waves traveling faster than those that are verticallypolarized, is present under continental cratons at 250-400 km depths and under oceanic plates atshallower (80-250 km) depth (Gung et al., 2003). This anisotropy has been related to horizontalshear in the low-viscosity asthenospheric channel, which is thinner below the continents thanbeneath the oceans (Gung et al., 2003). This is in agreement with a shear in the asthenospheredistributed worldwide. A global shear wave splitting analysis in the asthenosphere (Debayle etal., 2005) show directions consistent with a mantle shear along the undulating pattern of flowsuggested by surficial plate motions (Doglioni, 1993). Deviations from this flow occur particu-larly along subduction zones where the flow is inferred to encroach the slabs.

    Horizontal plate speeds range between 1-150 mm/yr, whereas vertical motion (uplift orsubsidence) of the lithosphere typically has rates between 0.01-1 mm/yr. Therefore, on average,horizontal velocities are 10 to 100 times larger, suggesting the greater importance of tangentialforces acting on the lithosphere, i.e. the toroidal field.

    The westward drift can be interpreted as a toroidal field of degree one (Ricard et al., 1991).Bokelmann (2002) suggests that in order to explain the toroidal component, plates and mantlecannot be fully coupled. He also proposed that the mantle is the dominant force in movingNorth America, based on the dip and orientation of P-wave fast azimuths axes. Holtzman et al.(2003), modeling a decoupling zone, showed that in simple shear experiments on several mantle-like melt-rock systems at high temperature and pressure, melt segregates into distinct melt-richlayers oriented 20 to the shear plane. As an application, in real peridotites, melt-rich bandsdipping toward the sense of motion can develop in the mantle near a shear zone. Accordingto the dip of the fast azimuths axes described by Bokelmann (2002), this would indicate thatmantle beneath Western North America can move eastward relative to the lithosphere, as alsosuggested by Silver and Holt (2004). Similar eastward asthenospheric flow has been interpretedby Negredo et al. (2004) for the Caribbean plate. Upper mantle anisotropy parallel to platemotion suggests eastward asthenospheric flow even beneath the East Pacific Rise (Wolfe andSolomon, 1998).

    Horizontal shearing across the asthenospheric decoupling zone implies that a force is acting

    3

  • on the lithosphere that opposes to di